Drug Safety Assessment in the Era of ICT Advancement

Transcription

1 20th DIA Annual Workshop for Clinical Data Management Forerunners History of 20 years and looking forward the future February 6-7, 2017 Tower Hall Funabori Drug Safety Assessment in the Era of ICT Advancement Dr Yoshiaki Uyama Director Office of Medical Informatics and Epidemiology Pharmaceuticals and Medical Devices Agency (PMDA)

2 Disclaimer The views and opinions expressed in the following PowerPoint slides are those of the individual presenter and should not be attributed to DIA, its directors, officers, employees, volunteers, members, chapters, councils, Communities or affiliates, or any organization with which the presenter is employed or affiliated. These PowerPoint slides are the intellectual property of the individual presenter and are protected under the copyright laws of the United States of America and other countries. Used by permission. All rights reserved. DIA and the DIA logo are registered trademarks or trademarks of Drug Information Association Inc. All other trademarks are the property of their respective owners DIA, Inc. All rights reserved. 2

3 Priority area in PMDA Nature Reviews Drug Discovery 13, 490 (2014) 1.More scientific contributions during development through consultation 2.Utilizing BIG DATA for improving quality of approval review and safety assessment 3.Promoting regulatory science Developing methods and criteria for responding to advances in science and more 2017 DIA, Inc. All rights reserved. 3

4 Limitations of available data at the approval Too Few:small sample size Too Simple:strict dosage and administration Too brief:shorter period Too median-aged:age limitation Too Narrow:excluding special population Rogers A.S., Drug Intel Clin Pharm., 1987 e.g.; Few elderly populations Hypertension Clinical Trial Clinical Practice 2017 DIA, Inc. All rights reserved. Asahina, Y. et al (2014) J Nutr Health Aging 18(5):

5 Rofecoxib and Cardiovascular risk Cardiovascular Events Associated with Rofecoxib in a Colorectal Adenoma Chemoprevention Trial (APPROVe Trial) Robert S et al, N Engl J Med, 352: , DIA, Inc. All rights reserved. 5

6 Earlier detection of risk signal by analyzing electronic health records Brown JS et al. Pharmacoepi Drug Saf, 16: , DIA, Inc. All rights reserved. 6

7 Pharmacoepidemiological activities in all over the world Europe US MID-NET Asian countries 2017 DIA, Inc. All rights reserved. 7

8 MIHARI Project To establish a framework of quantitative evaluation by pharmacoepidemiological methods using electronic healthcare data in PMDA Drug safety assessment using conventional information sources Spontaneous ADR report DB Overseas Literature regulatory actions PMDA Drug safety assessment using electronic healthcare data Claims DB MID-NET (EMR DB) MHLW Safety measure Medical institutions Risk communi cation Presentation in Academic Conference etc DPC DB 2017 DIA, Inc. All rights reserved. 8

9 Examples in MIAHRI project 2017 DIA, Inc. All rights reserved. 9

10 Various database from different sources Electronic medical records database Dispensing claims database Healthcare Delivery Medical Institutions Dispensing Pharmacy Claims Copayment Payment Agency Subscriber Examination payment facility Insurance fee Payment Issue Insurance cards Claims Insurer health insurance association s claims database MHLW National Claims Database 2017 DIA, Inc. All rights reserved. 10

11 Number of patients(adolescent) Example: Impacts of regulatory action (anti-influenza drug) Objectives: To assess impacts of regulatory safety measure for an individual products using the Japanese claims data Oseltamivir Zanamivir Approval of Pediatric indication to Zanamivir Label Change (Warning) Oseltamivir in Pediatric and Adolescent Label Change (Warning) Zanamivir in Pediatric and Adolescent Year/Month 2017 DIA, Inc. All rights reserved. 11

12 Risk of Acute Asthma Attacks Associated With NSAIDs: A Self-Controlled Case Series. observation start Baseline Period R 1 R0 R2 R3 R4 NSAIDs start date NSAIDs end date R0 = 7 days before prescription start date R1 = the prescription start date R2 = 1 9 days after the prescription start date R3 = > 9 days after the prescription start date R4 = 7 days after the prescription end date observation end Baseline Period Definition of acute asthma attacks: the combination of an inhalation procedure and the prescription of any inhaled β2-agonist. Takeuchi Y et al. Therapeut Innov Reg Sci, DIA, Inc. All rights reserved. 12

13 Risk of Acute Asthma Attacks Associated With NSAIDs: A Self-Controlled Case Series. Takeuchi Y et al. Therapeut Innov Reg Sci, DIA, Inc. All rights reserved. 13

14 MIHARI Example 2: Cardiovascular risks associated with Dipeptidyl-Pepridase 4 inhibitors: a cohort study Health insurance association s database Pharmacoepidemiology and Drug Safety; 24.SUPPL1.(529) Diabetes patient BG Outcome Acute myocardial infarction Cardiac failure Cerebral infarction Intracerebral hemorrhage Subarachnoid hemorrhage DPP-IV Estimated HR of CV risks in DPP4Is group compared with BGs Exposure No. of Adjusted HR Cases (95%CI) BG 25 1 DPP IV (0.67, 1.82 ) BG 61 1 DPP IV (0.75, 1.42) BG 23 1 DPP IV (0.55, 1.52) BG 6 1 DPP IV ( ) BG 0 - DPP IV 5 Not available DIA, Inc. All rights reserved. 14

15 MIHARI Example 3: Risk evaluation of Atypical Antipsychotics (AAP) for Hyperlipidemia: A Sequence Symmetry Analysis Risperidone Olanzapine Aripiprazole Adjusted sequence ratio (95 % CI) Takeuchi Y et al, Drug Saf (2015)38: Number of days since an initial administration of AAP 2017 DIA, Inc. All rights reserved. 15

16 MID-NET project 2017 DIA, Inc. All rights reserved. 16

17 MID-NET Project The Medical Information Database Network in Japan for a real-time assessment of drug safety (currently 4M patients) PMDA Hospital Database HIS data Claims data DPC data Hospital DB Hospital Hospital DB 23 hospitals In closed network Hospital DB DB 2017 DIA, Inc. All rights reserved. An integrated real time EMR database with high quality 17

18 Data categories in the MID-NET system Database HIS data Claims data DPC data HIS data Patient identifying data Medical examination history data (including admission, discharge data) Disease order data Discharge summary data Prescription order/compiled data Injection order/compiled data Laboratory test data Radiographic inspection data Physiological laboratory data Therapeutic drug monitoring data Bacteriological test data 2017 DIA, Inc. All rights reserved. 18

19 Onsite Center Overview of MID-NET Central data center User Summarized data 9 Send only summarized data (not individual data) 8 Output Summarized data individual level data 1Create program SAS etc 7View & Analysis SAS etc Summarized data Hospitals 2Request for running program 5Approve to send data 6Send data Technical staff for MID-NET Original databases Medical record Claims Labo test data Others 3Approve the request Standardization Anonymizition SAS etc Common data model database for MID-NET 4 Output individual level data OR Summarized data 2017 DIA, Inc. All rights reserved. 19

20 EMR Data Quality of MID-NET PMDA has worked with cooperative hospitals Before quality management for assuring data quality of MID-NET MID-NET EMR After quality management MID-NET compare compare Disease order data 99.1% Prescription order data 67.0% 品質管理作業後 Disease order data 99.9% Prescription order data 100% Laboratory test data 55.8% Laboratory test data 100.0% 2017 DIA, Inc. All rights reserved. 20

21 Incidence proportion MID-NET pilot: Ranmark (Denosumab) Hypcalcemia Launched ( ) Spontaneous ADR reports Serious Hypocalcemia: 32 cases Death:2 cases (~ ) Traditional process Dear healthcare professionals letter ( ) If we utilize MID-NET Quantitative risk assessment compared with control 0.6 Number of patients Number of patient with Hypocalcemia Incidence proportion Relative risk Denosumab Zoledronic acid Data from 3 hospitals(2013/7~12) 2017 DIA, Inc. All rights reserved. 21

22 (%) MID-NET pilot: Prazaxa (dabigatran) GI-Bleeding Compare risk of GI bleeding between Prazaxa and Warfarin Number of Prescription Results from 1 cooperative hospital of MID-NET Number of Patients GI Bleeding Patient number Prazaxa % warfarin 14,534 1, % % prazaxa warfarin Patients distribution based on Cr at the time of first prescription Results from 1 cooperative hospital of MID-NET Number of patients Normal -0.9mg/dL Mild mg/dL Moderate mg/dL Severe 2.7-mg/dL No Lab-test n % n % n % n % n % prazaxa % % 7 4.3% 0 0.0% % warfarin 1, % % % % % 2017 DIA, Inc. All rights reserved. 22

23 MID-NET pilot: Harvoni (Ledipasvir&Sofosbuvir) SUV in HCV Comparison of SUV with Daklinza MID-NET Hospital A (15 M data) Prescri ptions Patients Patient with Clinical examination HCV-RNA Patient % a with SUV b Harvoni % % % c Harvoni Daklinza Comparison of renal dysfunction sevirality with Daklinza Patients -0.9 mg/dl (Normal) Daklinza % % mg/dl mg/dl a) Percent to Patient b) SUV: Sustained virological response c) Percent to Patient with clinical examination 2.7- mg/dl No labo test N % N % N % N % N % Harvoni % % 1 0.4% 0 0.0% % Daklinza % % 0 0.0% 1 1.4% % 2017 DIA, Inc. All rights reserved. 23

24 Points to consider in utilizing EMRs for drug evaluation -PMDA s experiences in MIHARI DIA, Inc. All rights reserved. 24

25 Database reliability In addition to ensure an appropriateness of study design and data analysis, quality of database itself should be checked in advance. Selection of database Characteristics of database (data holder, data periods, sample size, patients background, traceability, collected items, procedure for access etc.) should be confirmed in advance 2017 DIA, Inc. All rights reserved. 25

26 Proper planning and design of a study and analysis Refer to the guideline on conduct of pharmacoepidemiological study utilizing medical record database for drug safety assessment(published on March 31 st, 2016) Make all efforts to understand how a target item was used in clinical practice Different diagnosis for claim Carefully consider clinical meaningfulness of an event definition Set a comparator for better interpretation of results 2017 DIA, Inc. All rights reserved. 26

27 Selection of appropriate data period and timing for a study Generally, data for a few years A timing for a study how many years after approval would be appropriate for a study purpose? An integrated assessment based on results of more than one study Confirm in 2 or more studies Careful assessment with consideration of study limitations 2017 DIA, Inc. All rights reserved. 27

28 Patients Registries 2017 DIA, Inc. All rights reserved. 28

29 Patients registries and Challenges Current registries mainly focus on academic research for grasping patient backgrounds in clinical practice A lack of items for regulatory review Less interventional data (e.g.: prescription date, drug name, dose, prescription period etc.) Less standardized data Non-coded data A lack of quality management Small sample size A patient registries which can be utilized for post-marketing study of a new drug and/or for drug development Establishment of Clinical Innovation Network (CIN) 2017 DIA, Inc. All rights reserved. 29

30 Clinical Innovation Network and PMDA MHLW PMDA CIN-Working Group Advice, Cooperation AMED About 20 members from New drugs & Safety Offices Muscular dystrophy Registry by NCNP ALS (Antilymphocytic serum) Registry By Nagoya Univ. Cancer registry By National Cancer Center Japan Cerebral surgery By Japan neurosurgical society Study group for epidemiological methods and data quality standards Study group for ethical issues for registries and relationships with industries Output Utilizing registry data for promoting cost effective clinical studies, for accelerating drug development, and for extending healthy life expectancy 2017 DIA, Inc. All rights reserved. 30

31 Recent PMDA initiative Big Data-utilized Assessment and Regulation 2017 DIA, Inc. All rights reserved. 31

32 CDISC data submission on NDA formally started on October 1st, 2016 NDA etc. CDISC data Database of Clinical Trial Results Analysis Modeling & Simulation: Concentration-Response Model, PBPK: Physiologically-based Pharmacokinetic Model etc. NDA Review B/R evaluation with raw data analysis Scientific Consultation Scientific advices based on the information obtained from analyses including M&S Cross-Products Analysis More effective & High quality review More efficient & Successful development More evidences & Advancing Regulatory Science Establish disease models Identifying common risk factors among different drugs etc DIA, Inc. All rights reserved. 32

33 PMDA Regulatory Science Center (planned in 2018) Regulatory Science Center BIG DATA -utilized Assessment & Regulation Archives of e-data EMR Data CDISC Data Active Utilization Utilization of e-data for better regulatory decision in Development Pre-Approval Pharmacovigilance Better B/R balance 2017 DIA, Inc. All rights reserved. 33

34 Future Model 2017 DIA, Inc. All rights reserved. 34

35 New technology for assessment Examples Wearable devices Smart watch etc. AI(artificial intelligence) Watson etc. More electronic data will be available and be accumulated!! 2017 DIA, Inc. All rights reserved. 35

36 Private information Organization for managing anonymized medical information on data utilization Hospital A Hospital B Hospital C 医療情報匿名 加工提供機関 ( 仮称 ) Organization for managing anonymized medical information on data utilization Anonymized data for analysis PMDA Institutions Industries Data User Review Safety Measures etc. Research etc. Drug development Pharmacovigillance etc. Utilization of real-world data for promoting public health DIA, Inc. All rights reserved. 36

37 Challenges Avoid being confounded by information Promoting standardization of data format for integrated assessments Establish standardized methods for utilization/analysis of the data/information [Important] International cooperation for sharing experiences/ knowledge in utilizing those data for regulatory purposes 2017 DIA, Inc. All rights reserved. 37

38 MHLW/PMDA Industries Medical Institutions Active utilization of EHR databases toward advanced medical care Regulatory decisions based on better scientific evidences Proper safety assessment utilizing HER databases in addition to the traditional approaches RMP implementation utilizing EHR databases Efficient risk management Better quality of safety information Provide leading-edge Medical Therapy with ensuring Safety Scientific and speedy safety measure Public 2017 DIA, Inc. All rights reserved. Better quality of Medical Care Maximize benefit/risk ratio 38

39 Regulatory Science Bridge Regulatory Science Bridge Tominaga T et al, Clin Pharmacol Ther, 90: 29-31, 2011 Data assessment Products of science (Substance, Knowledge, Information) Patients/ Society Stronger & More Complete Regulatory Science Bridge will help us in the future drug developments Tominaga T et al, Clin Pharmacol Ther, 90: 29-31, DIA, Inc. All rights reserved. 39

40 Ask